Use the confidence intervals generated by the Fisher method to determine whether two means are different:

·    If an interval does not contain zero, there is a statistically significant difference between the corresponding means.

·    If the interval does contain zero, the difference between the means is not statistically significant.

The individual confidence level indicates the percentage of times that a confidence interval contains the true difference for one comparison if the study were repeated multiple times.

For the Fisher method, examine the simultaneous confidence level because it can decrease to an unacceptable level. To display the simultaneous confidence level and the values of the confidence limits in the Session window, check Tests and confidence intervals in Stat > ANOVA > General Linear Model > Comparisons > Results.

For the salary analysis, all pairwise comparisons were requested for the subject factor. The confidence level chosen for the intervals was 95%, which corresponds to an individual error rate of 0.05 (or 5%). Because there are four levels of subject, this produces six pairwise comparisons. The confidence intervals for the comparisons reveal the following:

·    The confidence intervals for the differences between the means for subject 1, and the mean for subjects 2, 3, and 4 all contain only values greater than zero. This indicates that all other means are significantly greater than the mean of teaching subject 1.

·    The confidence interval for the difference between the means for subjects 2 and 3, also contains only positive numbers, which indicates that the mean for subject 3 is significantly greater than that for subject 2.

·    The confidence interval for the difference between the means for subjects 2 and 4 contains zero, which indicates that this difference is not significant.

·    The confidence interval for the difference between the means of subjects 3 and 4 contains only negative numbers, which indicates that the mean for subject 4 is significantly lower than that for subject 3.

The hypothesis test table indicates that the simultaneous confidence level for this family of comparisons is 80.38%. We can be 80.38% confident that all of these confidence intervals contain the true differences.